From: Dan Johnson <computerdruid@gmail.com>
Date: Wed, 25 Jul 2012 20:43:38 GMT
Subject: Re: False positive from orphaned_commit_warning() ?
Message-ID: <CAPBPrnugVm0RS5+Ljgg2E-AJygYsOSRjZW0Z=o9Mavs-SxTBog@mail.gmail.com>
In-Reply-To: <20120725185343.GA6937@windriver.com>

On Wed, Jul 25, 2012 at 2:53 PM, Paul Gortmaker
<paul.gortmaker@windriver.com> wrote:
> Has anyone else noticed false positives coming from the
> orphan check?  It is warning me about commits that are
> clearly on master.  Here is an example, where I checkout
> master~2 and then switch back to master.  It somehow thinks
> that master~2 is orphaned, when master~2 is by definition
> in the commit chain leading to master.

I've been able to reproduce this with the following simplified recipe,
although I still don't know what is causing the failure (I'm not very
familiar with the code)

git init test
cd test
#make 3 commits
touch a && git add a && git commit -m a
touch b && git add b && git commit -m b
touch c && git add c && git commit -m c

#clone it
cd ..
git clone test test2
cd test2
git checkout master~2
git checkout master
#Warning: you are leaving 1 commit behind, not connected to
#any of your branches


I can't figure out what's going wrong here, but the clone is
important; it doesn't fail without it. It appears to have something to
do with the fact that the cloned repository has a remote, as:
#in test2
git remote rm origin
git checkout master~2
git checkout master

Does not throw the warning, but it's not just the presence of
origin/master that triggers it, as:

cd ../test
git remote add origin ../test2
git fetch origin
git checkout master~2
git checkout master

Does not trigger it either.

Confused,
-- 
-Dan